The characteristics of the intervention and control group samples have now been analysed in order to check that the stratified random sampling that was carried out had resulted in two groups being constructed that were not significantly different from each other with respect to vital measures of attainment and child’s characteristics at pre-test.
It was found that there were no significant differences between children in the two groups at pre-test on gender distribution or results from the British Ability Scales (BAS II). Analysis of the schools characteristics has also been carried out including the percentage of children receiving free-school means, data from ESTYN reports (NFER progress in Maths and English and SAT performance in English, Maths and Science at level 3 and above), ECERS-E and ECERS-ICT. It was found that there were no significant differences between the schools of the children in the two groups according to these demographics. We appear to have been successful in constructing a statistically valid ‘level playing field’ from which the intervention may be robustly evaluated.